AI Finds the Opening, but Relationships Win Deals: How Human-Centric Sellers Are Winning in an AI-Obsessed Sales World

For the last year or so, a certain strain of AI bravado has tried to make the future of selling sound obvious. One startup plastered billboards throughout New York City reading "Stop hiring humans!" and urging companies to hire its AI sales rep instead. To a commercial team already under pressure to do more with less, the message lands because it promises relief: fewer people, more coverage. The machine hunts, writes, follows up, logs the activity, and, if you believe the loudest version of the pitch, sells.

But seasoned enterprise sellers have seen too many hype cycles come and go to buy that story. Deb Kammer is the Chief Commercial Officer of Polaris I/O, an enterprise account intelligence platform, and she thinks a lot of companies are pointing the technology in exactly the wrong direction. If everyone has the same models, the same flood of external signals, and the same generative tools, then the intelligence itself stops being scarce. What remains scarce is judgment, timing, and the nerve to act on a signal before it hardens into an RFP. She argues against the reflex that the billboards are selling. "Sales is not about finding pain points," she says of the work that actually wins an account. "It's finding partnership points."

Kammer comes from the buyer's side where for more than two decades she has led go-to-market teams in enterprise telecom and cable for carriers like T-Mobile, and before she joined Polaris she ran its platform as a customer for over two years. It made her own team faster and more precise, she says, which is why she talks about the tool like an operator rather than a vendor. The software, in her telling, turns raw signal into an starting point that sellers can act on.

Same Data, Different Outcomes

"There's a difference between data and information and action," she says. "Everybody can have the same data. How do you transition that to information and how do you make that actionable to be valuable?" A model can tell a seller that a target hired a new operations chief or opened a new region. Every competitor's model can, too. The edge is whether a person can look at that and know what it changes for this customer, this week, and say it in a way that starts a conversation instead of adding to the noise.

Kammer describes her own mornings as triage. A couple hundred articles land in her inbox, and she works them like a jigsaw puzzle, sorting the corner pieces from the middle ones until they "serve up a picture of success." Then she sends a client a sentence, sometimes less, on why one development matters to them specifically. Ten people might get the same underlying summary, but each gets a line that means something only to them. "AI doesn't always do that," she says. That's precisely why it's valuable.

She's glad to give the machine the parts of the job she never wanted. She doesn't get paid, she says, to "spend all day logging activity into Salesforce," and that administrative residue is exactly what should be automated, so a person is freed for the relationship. Polaris has built the split into a role. Alongside the software sits a Commercial Insight Strategist, whose job is to read a signal "through the lens of the customer's business context" and turn it into an account play a team can actually run.

Klarna's Cautionary Tale

The market is starting to relearn that division the expensive way. Fintech giant Klarna became the showcase for the efficiency case when it said its AI assistant was doing the equivalent work of 700 full-time agents and handling two thirds of its customer service chats. About a year later its chief executive, Sebastian Siemiatkowski, told Bloomberg that cost had been "a too predominant evaluation factor," and that "what you end up having is lower quality."

But the company didn't abandon the AI. It kept scaling it, later reporting the assistant could do the work of more than 853 agents even as it moved people back to the work that needed them. Volume to the machine, relationship to the person.

The Work Up Close

For Kammer, that work looks less like a billboard and more like homework. The way into a hard account is rarely a cold pitch, but a development the buyer already cares about. She describes finding, in a prospect's own investor call, a public commitment to sustainability, then working down through the organization to the ten or so people whose jobs that commitment touched, and spelling out why it mattered to each of them. The signal was free and public. The relevance was built by hand. "Every person matters and every person has a stake in the game," she says, and the job was making each of them feel it.

"If you are just a transactional salesperson, none of this matters," says Kammer. The signals and strategists approach assumes you're trying to build something with a customer rather than close them and move on. It's the same reason she stays a little wary of the tools she sells around. "Is it truthful? Is it telling me what I want to hear?" she asks of any AI output. A model can surface the signal. Only a person can test it against what they actually know about the human on the other end. Kammer is glad to hand the machine everything it can carry, and the sellers who last will do the same: learn the tools, give the systems the chores, and put what they save back into the relationship.